From: kaf@nwlink.com
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Neill Kapron <nkapron@google.com>, linux-usb@vger.kernel.org
Subject: Re: [PATCH] usb: gadget: u_audio: don't exceed maxpacket when adding a residue frame
Date: Mon, 16 Nov 2026 13:01:53 -0800 [thread overview]
Message-ID: <db18fcc4f778afbf515fed81a4a13488@nwlink.com> (raw)
In-Reply-To: <20261026231026.721536-1-kaf@nwlink.com>
On 2026-10-26 16:10, Kyle Farrell wrote:
> u_audio_iso_complete() clamps the base packet size to ep->maxpacket
> and accumulates fractional-frame residue in p_residue_mil. Once a
> whole frame has accumulated, it is appended to the packet without
> checking whether it fits within ep->maxpacket.
>
> Only append the extra frame if it fits. Consume the accumulated
> residue whether or not the frame is sent, preventing it from growing
> when the additional frame cannot fit. When the frame fits, behavior
> is unchanged. Otherwise, the stream cannot sustain the requested
> sample rate.
>
> For example, with 27 channels of 32-bit samples at 19200 Hz on a
> high-speed link with p_hs_bint=3, the required packet size of 1080
> bytes is capped at 1024. The base packet is 972 bytes, but appending
> a residue frame increases it to 1080 bytes, exceeding the allocated
> request buffer by 56 bytes.
>
> Fixes: 9bb87f168931 ("usb: gadget: f_uac2: send reasonably sized
> packets")
> Link:
> https://lore.kernel.org/linux-usb/sashiko-outbox-174688@kernel.org/
> Signed-off-by: Kyle Farrell <kaf@nwlink.com>
> ---
> drivers/usb/gadget/function/u_audio.c | 8 ++++++--
> 1 file changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/usb/gadget/function/u_audio.c
> b/drivers/usb/gadget/function/u_audio.c
> index f0ee83abcebc..f73d1085c1d1 100644
> --- a/drivers/usb/gadget/function/u_audio.c
> +++ b/drivers/usb/gadget/function/u_audio.c
> @@ -230,9 +230,13 @@ static void u_audio_iso_complete(struct usb_ep
> *ep, struct usb_request *req)
> do_div(div_result, uac->p_interval);
> do_div(div_result, 1000000);
> if ((unsigned int) div_result >= uac->p_framesize) {
> - req->length += uac->p_framesize;
> + /* req->length is bounded by ep->maxpacket above. */
> + if (uac->p_framesize <= ep->maxpacket - req->length) {
> + req->length += uac->p_framesize;
> + pr_debug("increased req length to %d\n", req->length);
> + }
> +
> uac->p_residue_mil -= uac->p_framesize * p_interval_mil;
> - pr_debug("increased req length to %d\n", req->length);
> }
> pr_debug("remains uac->p_residue_mil %llu\n", uac->p_residue_mil);
>
>
> base-commit: bdf12cde30830d46a8ef1ba93986356f1164b4f1
This patch also is suitable for stable backporting, as it fixes a
potential out-of-bounds write in existing configurations. Please add
the following trailer when applying:
Cc: stable@vger.kernel.org
Thanks,
Kyle Farrell
prev parent reply other threads:[~2026-11-16 21:01 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-26 23:10 [PATCH] usb: gadget: u_audio: don't exceed maxpacket when adding a residue frame Kyle Farrell
2026-10-26 23:15 ` sashiko-bot
2026-11-16 21:01 ` kaf [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=db18fcc4f778afbf515fed81a4a13488@nwlink.com \
--to=kaf@nwlink.com \
--cc=gregkh@linuxfoundation.org \
--cc=linux-usb@vger.kernel.org \
--cc=nkapron@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox